Each time there is an interaction with the environment, we select one among many, equally possible, actions. How is it done? In the classical view, to access motor resources, many cortical action plans compete in the basal ganglia: a set of subcortical neuron populations, whose output constantly inhibits specific thalamic targets. In the basal ganglia, one cortical plan is selected, the corresponding thalamic inhibition is disabled, finally producing the action (Albin et al., 1989). The classical view of action selection is now extended, including thalamic afferents. And, in the current view, a key role as modulators of action selection has been assigned to cholinergic interneurons of the striatum, the largest basal ganglia structure (Ding et al., 2010). Striatum is divided in two major populations of GABAergic medium spiny neurons (MSNs) according to different expression of D1 and D2 dopamine receptors. These two populations and their downstream pathways are respectively addressed as direct and indirect, for their connections with output basal structures (Tepper and Bolam, 2004). More than 90% of striatal neurons are MSNs while the remaining 5–10% are two types of interneurons, cholinergic, and GABAergic. These are providing the lateral inhibition that can actually implement the competition among cortical plans and therefore are a key to understand action selection. Cholinergic interneurons are known as tonically active neurons (TANs) for their spontaneous activity (>4 Hz). They receive glutamatergic afferents from cortex and thalamus, forming more synapses with thalamic than cortical terminals (Doig et al., 2014), and dopaminergic afferents from substantia nigra pars compacta. Locally, they branch dendritic arborization up to a millimeter in diameter, and they innervate GABAergic interneurons, other TANs and MSNs (Tepper and Bolam, 2004). Under resting conditions, MSNs are hyperpolarized by TANs spontaneous tonic firing (Tepper and Bolam, 2004). Morris et al. (2004) showed that increased cortical activity makes TANs respond with a brief burst, an afterhyperpolarization pause in their firing, and a rebound burst. In addition, Ding et al. (2010) have shown, in vitro, that also thalamic stimuli elicit TAN burst-pause pattern response. The work of Doig et al. (2014) aims at functionally distinguish thalamic and cortical contributions to action selection. Using different stimuli (single-, paired-pulse, and pulse trains) selectively delivered in the cortex and thalamus of rats, they were able to identify two temporal TAN response patterns. As in Morris et al. (2004), they used peri-stimulus time histogram (PSTH) analysis to identify the three main phases in TAN response: burst, pause and rebound. Applying pulse trains, which mimic the timing of motivationally salient stimuli (Ding et al., 2010), they were able to discriminate cortical and thalamic stimuli, predicting the whole TAN response based only on the initial phase. In fact, the number of spikes in the burst correlated positively with pause length and negatively with the magnitude of rebound: thalamic source was identified by high initial phase, long pause and low rebound, while cortical source was marked by lower initial phase, short pause, and higher rebound (Figure 1A).